This presentation will address two aspects of the usual approach used to establish the efficacy of an experimental treatment on the basis of a noninferiority trial. First, the concept of setting a non-inferiority margin is routinely used and is recommended in various regulatory guidance documents; however, it is extremely inefficient relative to an approach which pools data from the non-inferiority trial with historical data comparing the active control with placebo. Second, there is the concept that an experimental treatment must demonstrate preservation of the active control’s effect. While this appears on its surface to be a reasonable requirement, on closer inspection it can be shown to lead to serious logical inconsistencies. In this presentation I will propose a new approach for establishing efficacy that avoids the inconsistencies and inefficiencies of our current approach.